Upload
others
View
12
Download
0
Embed Size (px)
Citation preview
Nonreductive Physicalism or Emergent Dualism?
The Argument from Mental Causation
Timothy O'Connor and John Ross Churchill
Throughout the 1990s, Jaegwon Kim developed a line of argument that what purport to
be nonreductive forms of physicalism are ultimately untenable, since they cannot
accommodate the causal efficacy of mental states. His argument has received a great deal
of discussion, much of it critical. We believe that, while the argument needs some
tweaking, its basic thrust is sound. In what follows, we will lay out our preferred version
of the argument and highlight its essential dependence on a causal-powers metaphysic, a
dependence that Kim does not acknowledge in his official presentations of the argument.i
We then discuss two recent physicalist strategies for preserving the causal efficacy of the
mental in the face of this sort of challenge, strategies that (ostensibly) endorse a causal
powers metaphysics of properties while offering distinctive accounts of the physical
realization of mental properties. We argue that neither picture can be satisfactorily
worked out, and that seeing why they fail strongly suggests that nonreductive physicalism
and a causal powers metaphysic are not compatible, as our original argument contends.
Since we also believe that robust realism concerning mental causation should not be
abandoned, we take the argument of this paper to strongly motivate an account on which
the mental is unrealized by and ontologically emergent from the physical. In a final
section, we sketch what an ontologically emergentist account of the mental might look
like.
2
A Causal Powers Ontology
Let us first explain what we mean by the term “causal powers.” One way of using this
term is merely a loose manner of describing the causal features of a property, entity, or
kind independently of any definite commitments on the metaphysics of causation. A
person using the term this way might say, for example, that a defoliant has the causal
power to kill plants, where this claim is neutral as to whether (a) the presence of the
defoliant can produce or generate the death of a plant, (b) there is a law of nature that
relates properties of the defoliant and plant death, (c) plants regularly die after being
sprayed with the defoliant, (d) there are subjunctive conditionals relating the properties of
the defoliant and the death of plants in a certain way, (e) citing the presence of the
defoliant satisfactorily explains the occurrence of plant deaths in certain contexts, or
some (f) fitting a still different analysis of causation.
We do not use the term in this neutral manner. Our usage corresponds to the first
of these: a power to produce or to generate, where this is assumed to be a real relation
irreducible to more basic features of the world. Our favored technical term for this is
“causal oomph.” So understood, causation is not amenable to analysis in non-causal
terms, but instead involves the exercise of ontologically primitive causal powers or
capacities of particulars. Powers are either identical to, or figure into the identity
conditions of, certain of the object's properties, which are immanent to those things as
non-mereological parts. (Whether one thinks of these as immanent universals or tropes is
not crucial in this context.)
3
It bears emphasis that this view is not committed to assuming that all causation
must amount to something like “pushing,” or “pulling,” or “knocking,” or “the exertion
of a force.” What is assumed, rather, is solely this: when an instance of a property—the
event of the particular's having the property—is a cause, the world unfolds in a certain
way after the instance of that property, and that property instance is one of the factors that
jointly make the world unfold this way. This is just another way of saying what’s come
before, that the property instance, and others besides, jointly produce or generate certain
effects; they jointly oomph the world into going on in this way rather than that. Because
of this, there are certain counterfactuals true of the world (“were the property not to have
been instanced, such-and-such effects would not have occurred”). But these
counterfactuals are derivative from, and not to be equated with, or seen as the basis of,
the causal facts themselves: it’s because the property instance was among the factors that
jointly produced the relevant happenings that certain corresponding counterfactuals are
true. Causally efficacious properties have the power to make the world unfold in ways
that it otherwise would not, and this is a fundamental feature about these properties upon
which all else (counterfactuals true of them, regularities and patterns that encompass
them, explanations that cite them) is derivative.
There is much debate, and not a little confusion, over how to delineate the finer
points of this general picture. While we cannot delve deeply into these matters, we make
the following two remarks to forestall confusion that might infect understanding of our
subsequent argument. First, there is a pervasive manner of speaking that appears on the
surface to say that objects have and exercise causal powers. (Witness our example above
4
with respect to defoliants.) In our view, such talk should be construed by the causal
powers metaphysician as a shorthand way of expressing the claims that
i) the having of the property is the having of the causal power;
ii) the event of the property's being had by the object in appropriate circumstances
causally contributes to the effect; and
iii) the exercise of the causal power just is this causal contribution.ii
Second, a single property may contribute to a very wide array of effects, depending on
the context in which it is instanced. A particle’s being negatively charged may contribute
to its accelerating at varying rates away from a similarly charged nearby particle,
accelerating toward an oppositely charged nearby particle, even accelerating towards a
similarly charged particle (though at a slower rate than would occur were the particle not
to have been so charged), and countless other manifestations, all depending on the
context of its occurrence. But in ordinary speech, again, there is a tendency to talk of a
corresponding array of causal powers being exercised, ‘each’ of which is identified
through the effect actually manifested. This sort of speech has encouraged some
metaphysicians to posit a multiplicity of properties, or worse, to posit a distinct type of
entity (a causal power), any number of which are ‘conferred by’ a single property. We
should resist such moves on grounds of parsimony, and here science is a much better
guide to property/power identifications.iii The key is to understand a basic power or
disposition not in terms of this or that salient manifestation, but rather in terms of a
unitary causal influence, something that is constant across circumstances while its
manifestations will vary.
5
While we cannot undertake here a defense of this understanding of causation, we
will summarize what we take to be some key advantages over two very general rival
approaches. The first is the class of broadly Humean reductionist accounts. While details
differ considerably, on all Humean accounts, whether one event causes another is a
massively nonlocal matter, insofar as causal relations supervene on the global pattern of
events across space and time. Should the pattern of future events turn out to be very
different from what our best theories now predict, it might ‘turn out’ that what we
thought to be an obvious causal interaction—never mind the details of its nature, the very
existence of any causal relationship at all—was no such thing. But this is implausible.
What happens in the distant reaches of space and time cannot nullify whether a causal
transaction occurs here and now. (To focus, one’s intuitions, consider rational agency as a
special case. Actions are a special kind of causal process. If a Humean picture of
causation more generally is correct, then should the future pattern of events unfold in
certain ways, and nothing that has happened thus far will keep it from doing so, we
should have to say none of us ever undertook an action at all. Mind, we are not speaking
here of ‘free’ actions in the sense of free will, just plain-old actions.) The causal powers
metaphysician has no such implausible nonlocality commitment. Furthermore, by
grounding causal relationships, ultimately, in the contingent global distribution of
noncausal facts, certain kinds of explanations become unavailable in principle. For
example, if, in seeking an explanation for the occurrence of token event y, we’re seeking
knowledge of what made y occur, then the Humean must deny that there is any such item
of knowledge. And while, e.g., X- and Y-type events may conform to a pervasive pattern
of a specified sort (whether actual or counterfactual), for Humeans there will either be no
6
explanation as to why that pattern holds, or else the explanation will itself bottom out in
unexplained pattern facts. For anyone who shares our temperament vis-à-vis explanation,
these consequences are bound to disappoint.iv The account of causation we favor, on the
other hand, invites neither of these two disappointments. This should be obvious in the
first case, given what we have proposed concerning the causal relation. As for the second
case, noting that X-type events all manifest a common property that is disposed to bring
about Y-type events in specified sorts of circumstances provides us with an explanation
as to why X- and Y-type events conform to the pattern they do, an explanation that does
not bottom out in unexplained pattern facts. (And whatever else we might think of them,
Humean complaints that dispositions introduce a ‘mysterious modality’ into the world are
hardly founded in scientific practice, where functional methods of specifying theoretical
properties is standard. On the causal powers view, where such identifications are
accurate, they capture something about the nature of the property itself. Whereas for the
Humean, they merely describe contingent patterns of instantiation of the property, while
being forever silent concerning its intrinsic character.)
The second rival approach, developed in somewhat different ways by Michael
Tooley and David Armstrong, involves associating particular causal relations in the world
with a contingent, higher-order, nomic relation among universals, a relation that
Armstrong dubs ‘necessitation.’ The central idea is that the obtaining of a nomic fact of
the form N (F, G) grounds and explains the fact that a particular instance of F is followed
by a particular instance of G. But, as many have pointed out, it’s not at all transparent
how a second-order relation among universals can constrain particular first-order F-G
sequences.v Both Tooley and Armstrong have tried to complicate the account in order to
7
overcome the difficulty. Here we restrict our attention to Armstrong’s (1997) strategy. It
is to conjecture that the second-order relation among universals is identical to the causal
relation among its instances—causation is a relation not among particular, first-order
states of affairs, but of types of states of affairs. But now the appearance of an advantage
over the brute conjecture of the Humean theorist has vanished. For each occurrence of G
is ontologically and so explanatorily prior to the immanent, co-occurring N (F, G) fact.
So the posit of the N relation is gratuitous, as it can only be put into the world consequent
upon the regularity. It is merely a baroque adornment to Humeanism—enough so that we
might with justice call it “second-order Humeanism.”vi This is so, that is, unless we make
the stronger claim that F by its very nature is disposed to bring about G, in which case we
are back to the primitive dispositionality of the causal powers metaphysic.
Such, in outline form, are a few central reasons we have for thinking a causal
powers metaphysic to be preferable to its main rivals. In considering the prospects for a
nonreductive physicalist view of the mental, we are assuming, this metaphysic. We are
investigating its implications for the question at hand. Can the (by our lights) right-
thinking metaphysician who has seen his way clear to this view of causation make out a
nonreductive physicalist view on which mental states are causally efficacious in this
sense? We will try to persuade you that the prospects are bleak.
Causal Powers and the Dilemma of Reduction or Causal Exclusion of the Mental
We will now present our preferred version of a Kim-style argument for the reducibility of
mental properties to physical properties.vii We begin with three related premises
concerning causation and properties:
1) Causation is a real relation irreducible to more basic features of the world.
8
(causal nonreductionism)
2) Causation involves the exercise of causal powers or capacities of particulars.
(production account of causation)
3) Properties are individuated in terms of causal powers, such that there are no distinct
properties that confer exactly the same causal powers.
(causal theory of properties)
The next three premises flow from the distinctive commitments of non-reductive
physicalists:
4) Mental properties supervene on physical properties. (supervenience thesis)
The hoary slogan, of course, is ‘no mental difference without a physical difference,’
intended to capture an appropriate dependence relation. What exact form the
supervenience relation should take in this context, however, is a difficult and
controverted issue. We will follow Kim in supposing that complication arising, e.g., from
mental content externalism can be safely ignored. If this is correct, we may assume for
the sake of argument that mental properties 'strongly supervene' on the physical
properties of the individual (or on the physical properties and relations of the individual's
parts). Next we have:
5) Mental properties are realized by physical properties: a particular event M of a person
S’s having mental property M is either ‘constituted by’ (a kind of ontological
posteriority) or is identical to various physical particulars—possibly including portions of
the person’s environment—having certain physical properties and standing in certain
physical relations. (realization thesis)
9
We will be noncommittal on whether the realization of mental properties by physical
properties involves constitution or identity of the corresponding events, since non-
reductive physicalists’ pronouncements on this matter are varied and often obscure.viii
Finally, physicalists typically wish to assert:
6) Every physical event that has a cause has a complete physical cause.ix
(causal completeness of physics)
According to (6), nothing non-physical is required in order to causally account for the
occurrence of any physical event, where the latter consists in the instantiation of
fundamental physical properties and relations by fundamental physical entities. Whatever
their particular views concerning the status of special science laws and causes, including
those pertaining to psychology, the typical physicalist maintains that any fundamental
physical event (including large-scale distributions of fundamental properties and
relations) that has a cause has a cause that is equally fundamental and physical in
character. The true physics is causally complete.
We now contend that (1)-(6) are inconsistent with supposing
7) A mental property, M, is distinct from its physical realizer property (or properties), P,
and each event that consists of M’s being instanced exercises a distinctive form of
causality that one way or another impinges the realm of physical events.x (assumption for
reductio)
Premise (7) is the supposition that there are mental properties that do not reduce to
physical properties and whose causal efficacy does not reduce to the causal efficacy of
some physical properties. This means, in the schema used in (7), that the singular causal
10
action of the mental event of M’s being instanced does not reduce to the singular causal
action of some physical event or events (say, the instancing of the physical property P
that realizes M in the circumstances). In short, the commitment expressed by (7) is what
puts the “non” in “non-reductive” physicalism.
The argument that (7) is inconsistent with (1)-(6) proceeds as follows.
8) The instance of M either
(a) directly produces a subsequent mental event, M2, or
(b) it directly produces a wholly physical event, P*.
The realization thesis (5) and production account of causation (2) together strongly
suggest that option (a) is a nonstarter. On this view, mental events are ontologically
dependent on their subvening realizers, wholly constituted by (if not identical to) them,
and this is no less true of mental effects as of mental causes. Bringing about such a
mental event eo ipso involves causally affecting the physical event which realizes it. So
9) Not (8a).
But the thesis of causal completeness (6) implies that
10) If 8b, then the physical event P* is overdetermined by M and some other physical
event.
Now, if we accept the production account of causation, it will seem passing strange to
suppose that, in regular fashion, there are physical events that are systematically
“overoomphed” by distinct events, even if—indeed, especially if—these causes might
stand in a supervenience relation. If, say, a physical event P, the realizer of the mental
event M, produces or oomphs P*, what causal work is left over for M? It would be at best
11
a gross violation of parsimony to posit two distinct productive relations for a single event
every time mental events supervene on the fundamental physical cause. Note that on
reductive accounts of causation, on which causal facts are not something additional to the
totality of noncausal facts, the situation looks very different. Suppose, for example, that
our effect P* is counterfactually dependent on both P and M. If we accept something like
the counterfactual analysis of causation, there is nothing strange or objectionable about
deeming M, as well as P, to be a cause of P*. For in doing so we are not making a
commitment to anything additional—M’s status as a cause of P* falls out of the facts that
we already accept along with our analysis. It comes for free. By contrast, on the
nonreductive productive account, we would be positing an additional fundamental
relation between M and P*, when doing so is entirely unnecessary for accounting
causally for P*. Thus, we should conclude that
11) There is not systematic mental-physical overdetermination, as the consequent of 10
implies.
But this is the end of the road. We are forced to conclude, therefore, that
12) M does not make a distinctive contribution to occurrences in the physical world,
whether wholly physical or supervening mental occurrences. (Completing Reductio of 7.)
Finally, the causal theory of properties (premise 3) both rules out an epiphenomenalist
retreat and suggests the proper ultimate conclusion: we ought either to reductively
identify M with P or deny that M is a bona fide property—one that earns its causal
keep—in the first place.
12
Non-reductive physicalists see an obstacle to the first option, reductionism, in the
fact that, as functional properties, intentional properties are multiply realized. What
counts as a belief that Q in humans may be quite distinct, at any physical level of
description, from what counts as that same belief in, say, an intelligent extraterrestrial or
a sophisticated artificial machine built out of steel and silicon. In reply, Kim recommends
that we seek local, species-specific reductive identities for intentional properties—human
belief that such-and-such as identical with physical property so-and-so—and so preserve
the status of these intentional properties as causal powers. That is, we characterize both
M and P in terms of highly specific mental and physical types, respectively, and move to
a type-type identity theory.
The second, eliminativist option is to interpret apparent reference to mental
properties as properly denoting mental concepts only. There are far fewer properties had
by an object than the vast number of concepts it falls under. As indicated in premise (3),
properties are immanent to their instances and make a nonredundant difference to how
the objects act in at least some circumstances. (They answer to what Kim calls
'Alexander's dictum.')
The argument just presented, like earlier relatives, seeks a reductionist or
eliminativist conclusion by way of arguing for the exclusion of irreducibly mental
causation. Yet it does this by explicitly invoking the thesis of causal powers realistically
construed. So let us refer to it hereafter as the power exclusion argument.
As critics of Kim have observed, this argument appears to generalize beyond
mental properties to all properties posited in the special sciences (sciences other than
basic physics).xi And since, contra Kim, it is highly plausible that special science
13
categories are not ontologically reducible (owing in part to their own multiple
realizability)xii, the eliminativist conclusion the argument ultimately invites here is often
taken as a reductio ad absurdum: surely the terms of well-established biological and
chemical theory pick out genuinely efficacious properties!
Owing to length constraints, we shall not be able to treat this sort of indirect
criticism of the argument in detail. We will rest content with the following two-fold
response.
First, notice that a rejection of premises (5) and (6), the realization and causal
completeness theses, suffice to block the final conclusion of the power exclusion
argument. As we discuss later on, we believe the best way to maintain a robust,
nonreductive view of the mental is to reject these two premises. Similarly, one might
reject the corresponding premises in an exclusion argument directed at special science
properties that one takes to be irreducible and efficacious. Recent philosophy of science
has seen significant challenge to the completeness thesis in particular.xiii
But second, for one who takes the case for the completeness of physics with
respect to some or all special sciences to be convincing, it would not be absurd to accept
a causal exclusion conclusion from a corresponding form of argument. For so-called 'high
level' theories can be enormously useful and illuminating, and even necessary to the
progress of human knowledge of how the world works, without answering to ontological
'levels' or layers populated by distinctive properties and their objects.xiv And the further
fact that such theories are not generally reducible to more fundamental theories is a
highly interesting fact about our world (and one necessary for science to get off the
ground, as in practice we inevitably work our way in, not out), but it cuts no ontological
14
ice. An alternative to the levels picture of physical reality has already been hinted at
above: there is a vast array of microphysical entities (for simplicity, "the particles")
bearing primitive, dynamical features and standing in primitive relations. Talk of
composite objects and their properties, at least in the general case, is the imposition of a
conceptual scheme that selectively picks out coarse-grained patterns running through the
vast storm of particles. These concepts really are (objectively) satisfied by the world, but
not in virtue of a one-one relation between general concepts and properties, or individual
concepts and particulars.
This second response might be thought to entail the devaluation of the special
sciences. Such a conclusion would be too hasty, however. For it is simply false that
science is of value only as a source of representing the world in more and more accurate
ways. It is, in addition, a source of means for intervening and manipulating the world so
as to change it for the better, and much of its value is due to this rather than its
representational fruits. We value science—we fund it, prioritize it, give special social
status to many of its practitioners, etc.—because of its role in improving the world, and
not just because of its role in representing the world. (The development of methods for
effectively preventing and treating myriad diseases serves as just one example of such
improvement.) But qua sources of improvement, some of the special sciences are at least
as valuable, and perhaps more so, than fundamental physics. For we are very often better
able to intervene and manipulate in ways that improve the world by using the resources
of the non-fundamental special sciences.
Returning to the status of mental properties, here, then, is where we are left. The
commitments that drive the power exclusion argument—the causal powers metaphysic
15
along with the supervenience, realization, and causal completeness theses of nonreductive
physicalism—appear to generate the conclusion that mental properties are either
reducible or eliminable. This is serious trouble for philosophers that are neither
reductionists nor eliminativists with respect to the mental, and they are legion. Thus, if
we wish to preserve the mind as irreducibly efficacious, we must reject one or another of
the commitments driving the argument above.
Or so we believe. Sydney Shoemaker, however, disagrees, and has recently
attempted to provide a way out for the nonreductive physicalist who is a realist with
respect to causal powers. Since Shoemaker has bona fides as both a causal powers
metaphysician and as a physicalist, it is fitting that we investigate his approach in detail.
Shoemaker on Non-reductive Mental Causation
Shoemaker thinks that the key to vindicating the causal efficacy of mental properties
without reduction lies in a distinctive account of the realization of mental properties by
physical properties. In broad strokes, his proposal is that mental and other realized
properties are disjunctive properties, with their disjuncts as their realizers: the relation of
realizer to realized is simply the relation of disjunct to disjunction.xv This ensures that
realized properties have a proper subset of each of their realizers’ forward-looking causal
features—what instances of the properties can causally suffice for—while having a
superset of their realizer properties’ backward-looking causal features—what can
causally suffice for instances of the properties. Shoemaker then exploits the conclusion
that realized properties have a subset of their realizers’ powers to argue that mental
causation is not reducible to causation by the physical realizers, owing to a certain
16
proportionality thesis explained below concerning what counts as a cause of what. The
following schema captures Shoemaker’s picture.
(C1 v C2) is a property realized by each of C1 and C2 just in case: C1 —> E1 C1 —> (E1 v E2) B1 —> C1 B1 —> (C1 v C2) C2 —> E2 C2 —> (E1 v E2) B2 —> C2 B2 —> (C1 v C2) (C1 v C2) —> (E1 v E2) (B1 v B2) —> (C1 v C2) ("—>" denotes causal sufficiency)
How is accepting this picture of realization supposed to make things easier for
non-reductive physicalism? We begin by observing that if the realized property has a
subset of the forward-looking causal features of the realizer, then the realizer property
event is causally sufficient for everything the realized property event is causally sufficient
for, plus more. So, for example, C1 is causally sufficient for (E1 v E2), just as (C1 v C2)
is, but unlike the disjunctive cause it is also sufficient for an instance of E1. Now, if C1
and (C1 v C2) overlap in this way in what they causally suffice for, and if causal
considerations ought to drive our conclusions about the identity of properties, a natural
conclusion is that (C1 v C2) is a proper part of C1. More generally: events which
instance realized properties are parts of those instancing the corresponding realizers, and
so are not identical to them.
From here, Shoemaker invokes a version of Stephen Yablo’s “proportionality”
constraintxvi on what we ought to count as “the cause” in a causal interaction: while it is
true that C1 is causally sufficient for (E1 v E2), (C1 v C2) is, Yablo and Shoemaker say,
a better candidate for being the cause. For (C1 v C2) is also causally sufficient for the
specified effect, but only “just so”—it causally suffices for the effect and nothing more
besides. The only features of C1 that contribute to the “bringing about” of (E1 v E2) are
features had by (C1 v C2), a “part” of C1. And as with a more familiar sort of case, such
17
as Jones’s single shot in a firing squad, just ahead of the others, killing the condemned,
we are invited to conclude that it is best to say that while the whole (C1; the firing
squad’s firing) was causally sufficient for the effect ((E1 v E2); the death of the
condemned), proportionality constraints argue in favor of counting a particular part ((C1
v C2); Jones’s firing) rather than the whole as the cause.xvii This is how realized events
in general—and realized mental events in particular—qualify as causes in certain
scenarios, such that it is false in these scenarios that the (physical) realizer events are
likewise causes of the very same effects.
We now have Shoemaker’s account of realization laid out before us, as well as the
way it is supposed to provide for non-reductive mental causation. But how, we may
wonder, does the account underwrite a response to the power exclusion argument?
Notice, first, that there is no rejection of the supervenience, realization, or completeness
theses (premises 4-6). And, of course, Shoemaker intends (7), the anti-reduction premise,
to come out true as well. Thus, one who takes our exclusion argument above to be
cogent will naturally suspect that Shoemaker's commitment to the causal powers
metaphysics (as expressed in premises 1-3) is less than it appears.
To bring the problem into focus, consider first that, for all his distinctive claims,
Shoemaker clearly gives ontological priority to the physical realizer event. He tells us
that P realizes M just in case P is metaphysically sufficient for (but not identical to) M
and 'constitutively makes [it] real' (2007: 6, 10). He goes so far as to say that realized
states are 'nothing over and above' their realizers (2007: 2). If all this is so, then how is a
case of M's causing an effect, E, not also a case whereby P, M's constituting realizer, is
likewise causing E? Indeed, how is this not a case where P is causally prior to M, so that,
18
by the power exclusion argument, we should conclude that P is the sole true cause?xviii It
seems that only a retreat from a causal powers metaphysics could allow you to say that P
is somehow 'merely' causally sufficient whereas M is the proper cause. If P is
ontologically prior to M, able to bring about E, and in the circumstances necessary to do
so, how can it get out-oomphed by M?
The only way for us to make sense of this is by ignoring Shoemaker's talk of P's
constitutively making real M and focusing instead on his notion that M is a part of P,
owing to the subset-of-powers thesis and the causal theory of properties. (This line of
interpretation is encouraged by his invocation of the firing squad analogy (2007: 53) and
also by such statements as "It is only because the C-fiber stimulation instance realizer
contains the pain instance realizer that it has the relevant effects." (2007: 48)) We might
then suppose that Shoemaker is proposing what amounts to a radical inversion of the
reductionist's vision, such that it is the physical properties that resolve into an assemblage
of mental properties plus some non-mental remainder.xix Now, even if that could make
sense of how mental causation would be genuine, it has the substantial drawback that it is
just plain weird. The physical property is thereby conceived as (or as closely bound up
with) a cluster of causal features, a subset of which are the features that define an
associated mental property. Making this picture out perhaps requires us to analyze the
physical property as a structural property, the instancing of which just consists in the
instancing of properties of the object's parts and relations between them. The mental
property then comes out, on Shoemaker’s view, as an overlapping structural property,
perhaps somehow abstracted from the full physical structural property. Waiving
considerable doubts we should have about the plausibility of this picture of the mental-
19
physical relationship, it still has the result that both mental properties and the larger
structural physical properties in which they are embedded turn out to be derivative
structures, entities that are constructions out of microphysical properties and relations.
The specter of reductionism again menaces.
Shoemaker attempts to resist the reductive identification of mental (or other
macro-level) properties with microphysical states of affairs by giving an analysis of
microphysical realization on which there is only constitution, not identity. However, his
case for this rests on two claims about property identity that should be unacceptable to a
causal powers theorist (2007: 48-49). First, he lays down that, in general, a property
instance has just one constituent object and one constituent property, so a mental property
instance can't be identical to a state of affairs involving many distinct properties and
objects. He seems to put this forward as a definitional truth or platitude. But a causal
powers theorist does not take quasi-grammatical considerations to be final arbiters
concerning the structure of reality. One might just as well take Shoemaker's supposed
platitude together with facts (assumed for now) about microphysical constitution and
draw the conclusion that there are not, strictly speaking, mental properties at all—not in
the sense of entities that contribute directly to how the world unfolds. Shoemaker's
second claim is that the modal properties of macro-level property instances and their
microphysical realizers will generally differ. (Consider familiar claims made in
discussions of the statue of Goliath.) This claim rests on intuitive judgments about
possible variation in the material constitution of composite objects. But the status of
composite objects, no less than that of 'their' properties, is very much in question on the
powers metaphysics. One cannot simply assume that there are robustly objective modal
20
facts about them and use these to ward off what otherwise appears to be a powerful
reductionist challenge.
It is time to take stock. We have defended a power exclusion argument for the
untenability of nonreductive physicalism. It is a variant of Kim's argument that makes
explicit an assumption of a causal theory of properties. We have tried to show that
Sydney Shoemaker's recent attempt to harmonize the two positions fails. We believe that
such failure was inevitable, given Shoemaker's unusual combination of physicalist
commitments, a causal theory of properties, and an abundant, rather than sparse, ontology
of such properties. It is an unstable compound. The causal theory requires that properties
earn their keep. This appears inevitably to push the philosopher in one of two directions
when it comes to macroscopic structures: reduction or elimination, on the one hand (we
needn't here adjudicate the claims of these two), or a rejection of one or more of the
characteristic claims of physicalism, on the other.
A typical rejection of physicalism includes the denial of the realization, causal
completeness, and the supervenience premises. (We’ll here set aside the question whether
the rejection of physicalism requires rejecting all forms of supervenience, though we'll set
this issue aside, as it is not resolved easily and is not important for our purposes here
anyway.xx)
But before we consider what form a rejection of physicalism might take, we must
consider an alternative and original proposal defended in a number of places by Carl
Gillett. Gillett contends that we can best reject the Kim-style argument against non-
21
reductive physicalism not by rejecting physicalism itself, but by weakening it. He
suggests that we may retain realization and supervenience, and reject only completeness.
Gillett’s “Strong Emergence”
Gillett dubs his view "strong emergentism." On this view, mental properties, like all
macro-level properties, are realized microphysically. That is to say, an instance of a
mental property is identical to a combination of other, micro-physical property instances
and relations among them, where the other properties are, in the circumstances, the
property's realizers.xxi Strong emergence occurs where microphysical properties
contribute different fundamental causal powers to their microphysical individuals
precisely when these properties realize certain properties.
A schematic example: microphysical property L confers upon the microphysical
entity that bears it only powers α, β, and γ in all circumstances except when it realizes
property H, in which case it confers power δ. Now, notice that it’s still the case that only
microphysical properties contribute the fundamental, irreducible causal powers. (We’ll
hereafter let the qualifiers “fundamental” and “irreducible” be implicit.) But Gillett
argues that the realized property H is nevertheless causally efficacious, for three reasons:
i) H non-causally determines L to contribute the δ power to its bearer,
ii) H is a necessary member of a set of factors jointly sufficient for the contribution of
δ to an individual, and
iii) positing H as a causally efficacious property is necessary if we are to “account
for” the relevant microphysical individual’s having power δ.xxii
22
Furthermore, since in cases of strong emergence, we have realized properties, such as H,
determining the contribution of causal powers (like δ) by their micro-level realizer
properties (L), these count for Gillett as cases of “downward determination” by H.xxiii
Such, in brief, is Gillett's suggested picture of 'strong' emergence. How does it
fare as an explication of robust nonreductionism, and is it preferable to non-realizationist
varieties of emergence? Notice first that since H is a realized property, the microphysical
property L whose activity it non-causally determines is part of H itself. This may seem to
result in an objectionable circle of determination relations, but it does not. While an
instance of L contributes to the constitutive determination of an instance of H, the latter is
not thought to similarly bear some kind of ontological priority to the former. Instead, H
determines which causal powers L shall confer in the context.
Our basic criticism is this: Gillett's strong emergence provides at best a very
attenuated form of causal efficacy for mental properties. They do not produce (non-
derivatively) any event or even trigger some other causal power into activity. They seem
to be simply the occasion on which microphysical properties act in unusual ways (i.e.,
ways departing from their nearly ubiquitous manner of activity). In fact, from the
perspective of a causal powers theorist, H in our example seems but a handy name for the
sort of circumstances in which L confers δ; it answers one sort of “when” question. But
it’s hard to see what’s gained in explanation by insisting on accepting an emergent
realized property into our ontology. In fact, this insistence plausibly obscures, for (when
combined with “determination” talk) it makes it look as if there’s some light being shed
on how and why L confers δ when there’s not. The sober metaphysical truth seems to be
that whenever L is co-instanced with certain other properties in a certain way—where
23
that way and those other properties can be wholly specified in microphysical terms—then
L confers δ. We can introduce the term "H" as a label for this type of scenario. But in
doing so, we wouldn’t be accounting for anything that we hadn’t previously accounted
for in speaking only of microphysical properties and relations, and we wouldn’t have
gotten one step closer to understanding how or why L confers δ in the relevant scenarios.
Hence, we don’t think we ought to accept Gillett's contention that H, understood as an
emergent realized property, is a necessary posit in “accounting for” the contribution of δ
by L.xxiv
Some will be inclined to reply at this point along the lines of Jerry Fodor's brief
on behalf of the standard, non-emergentist variety of nonreductionism: we must
recognize H as a real, multiply realized, and explanatory property in its own right because
otherwise we will fail to capture the commonality of the many different scenarios,
microphysically described, in which L confers δ. Only here the case for H would be
bolstered by the fact (ex hypothesi) that the fundamental dynamics are distinctive in
nonreductive scenarios. We are unmoved, but suppose one is inclined to concede the
point. Even so, all we would have embraced are mental properties that play a kind of
structuring role in the world's dynamics. They do no distinctive causal work—provide no
extra causal oomph. There is, indeed, a strong analogy here to the role played by spatial
and temporal relations in Newtonian mechanics, as construed by a causal powers
theorist.xxv Such relations, one might say, provide a necessary framework for the
interplay of dispositional entities, while themselves having no dispositional nature. Surely
our nonreductionist physicalist wants more than this by way of the causal relevance of the
mental. More than being local, nondispositional constraints on the way fundamental
24
physical causes operate, our beliefs, desires, and intentions themselves directly contribute
to the unfolding dynamics of our behavior.
A Better Account of Emergence
It thus appears that a rejection of the causal completeness tenet of mainstream
physicalism will not in itself suffice to secure a robust efficacy for mental properties. We
must also reject the realization thesis, and in the context of mental causation, at least, that
is clearly a rejection of physicalism altogether. In our judgment, the best avenue for
developing an anti-physicalist view rooted in the rejection of realization and
completeness involves a stronger variety of emergence, what is often termed ontological
emergence.
The term "emergence" is used to cover a multitude of sympathies (in some cases,
sins). So we want to indicate in clear, albeit very abstract, terms what an emergentist
picture would look like, in our way of thinking.
Properties are ontologically emergent just in case:
(i) They are ontologically basic properties (token-distinct from, and unrealized by, any
structural properties of the system).
(ii) As basic properties, they constitute new powers in the systems that have them, powers
that non-redundantly contribute to the system’s collective causal power, which is
otherwise determined by the aggregations of, and relations between, the properties of
the system’s microphysical parts. Such non-redundant causal power necessarily
means a difference even at the microphysical level of the system’s unfolding
25
behavior. (This is compatible with the thesis that the laws of particle physics are
applicable to such systems. It requires only that such laws be supplemented to
account for the interaction of large-scale properties with the properties of small-scale
systems.)
In respects (i) and (ii), emergent properties are no less basic ontologically than unit
negative charge is taken to be by current physics. However, emergent and microphysical
properties differ in that
(iii) emergent properties appear in and only in organized complex systems of an
empirically specifiable sort and persist if and only if the system maintains the
requisite organized complexity. The sort of complexity at issue can be expected to be
insensitive to continuous small-scale dynamical changes at the microphysical
level.xxvi
We are inclined to further suppose (though this may depend on our inclination to accept a
controversial, strong causal explanatory principle) that
(iv) the appearance of emergent properties is causally originated and sustained by the
joint efficacy of the qualities and relations of some of the system’s fundamental parts.
(This would involve fundamental properties having latent dispositions to contribute to
effects, dispositions that are triggered only in organized complexes of the requisite
sort.)
26
One cannot give uncontroversial examples of emergent properties, of course.
Though there are ever so many macroscopic phenomena that seem to be governed by
principles of organization highly insensitive to microphysical dynamics, it remains an
open question whether such behavior is nonetheless wholly determined, in the final
analysis, by ordinary particle dynamics of microphysical structures in and around the
system in question.xxvii Given the intractable difficulties of trying to compute values for
the extremely large number of particles in any medium-sized system (as well as the
compounding error of innumerable applications of approximation techniques used even
in measuring small-scale systems), it may well forever be impossible in practice to
attempt to directly test for the presence or absence of a truly (ontologically) emergent
feature in a macroscopic system. Furthermore, it is difficult to try to spell out in any
detail the impact of such a property using a realistic (even if hypothetical) example, since
plausible candidates (e.g., phase state transitions or superconductivity in solid state
physics, protein functionality in biology, animal consciousness) would likely involve the
simultaneous emergence of multiple, interacting properties. Suffice it to say that if, for
example, a particular protein molecule were to have emergent properties, then the
unfolding dynamics of that molecule at a microscopic level would diverge in specifiable
ways from what an ideal particle physicist (lacking computational and precision
limitations) would expect by extrapolating from a complete understanding of the
dynamics of small-scale particle systems. The nature and degree of divergence would
provide a basis for capturing the distinctive contribution of the emergent features of the
molecule.
27
Now, many contemporary philosophers seem to think that such a view is too extreme
to be plausible. When pressed, such critics often cite the alleged consequence that an
emergentist view compromises the unity of nature. But unity does not require the
reductionist vision of the world as merely a vast network binding together local
microphysical facts, with a pervasive and uniform causal continuity underlying all
complex systems. It is enough that at every juncture introducing some new kind of
causally discontinuous behavior, there is a causal source for that discontinuity in the
network of dispositions that underlie it. In short: unity in the order of the unfolding
natural world need not involve causal continuity of behavior, only continuity of
dispositional structure.xxviii For the emergentist, the seeds of every emergent property and
the behavior it manifests are found within the world’s fundamental elements, in the form
of latent dispositions awaiting only the right context for manifestation.
We make no assertion one way or the other as to whether anything is like this for
any chemical or biological properties, though we note that present evidence allows for the
possibility that some perfectly respectable biological and chemical features are
ontologically emergent in this way.
We do, however, propose that the conscious intentional and phenomenal aspects of
the mind strongly favor an emergentist account. A human person’s experiences and other
conscious mental states exhibit features quite unlike those of physical objects, whether as
revealed in ordinary sense perception or as uncovered in the physical and biological
sciences. And the maximally direct nature of our first-person awareness of the intentional
and phenomenal features of our conscious states prohibits the a posteriori ascription to
them of underlying physical micro-structure hidden to introspection. The upshot of this
28
familiar reflection, if it stands, is that our experiences and other conscious mental states
have fundamentally distinctive characteristics. But these very characteristics are also
prima facie causally efficacious. (Indeed, on a causal powers metaphysic, to countenance
them as properties is to accept them as efficacious.) Thus, certain mental properties
appear to be (1) resistant to analysis in terms of physical structural properties and so
plausibly ontologically basic, (2) causally efficacious, and (3) borne only by highly
organized and complex systems. Though we cannot argue the matter at length here, we
find extant materialist attempts to overcome this prima facie case to be implausible.xxix (It
goes without saying that we take the grounds for an emergentist account of the mental to
be defeasible.)
Some philosophers acknowledge that the sort of broadly ‘Cartesian’ picture sketched
above captures how we naively think about conscious experience but contend that it is an
illusion. For our part, we think that such philosophers underestimate the difficulties for a
theory of empirical knowledge that maintains that we are subject to a radical and
pervasive cognitive illusion at the very source of all our empirical evidence. And if the
central argument of this paper is correct, then for any of these philosophers likewise
committed to a causal powers metaphysic, the seemingly paradoxical position of denying
the causal efficacy of mental states must be added to those difficulties.
Indiana University
29
Bibliography
Armstrong, David (1997). A World of States of Affairs. Cambridge: Cambridge
University Press.
Baker, Lynne Rudder (1993). "Metaphysics and Mental Causation," in Heil and Mele
1993: 75-95.
Block, Ned (1990). “Can the Mind Change the World?” in G. Boolos, ed., Meaning and
Method: Essays in Honor of Hilary Putnam. Cambridge: Cambridge University
Press, 137-170.
Block, Ned (1997). "Anti-Reductionism Slaps Back," Philosophical Perspectives 11:
Mind, Causation, and World: 107-132. Cambridge, MA: Blackwell.
Block, Ned (2003). "Do Causal Powers Drain Away?" Philosophy and Phenomenological
Research 67: 133-50.
Burge, Tyler (1993). "Mind-Body Causation and Explanatory Practice," in Heil and Mele
1993: 97-120.
Cartwright, Nancy (1983). How the Laws of Physics Lie. Oxford: Clarendon Press.
Cartwright, Nancy (1989). Nature’s Capacities and Their Measurement. Oxford:
Clarendon Press.
Cartwright, Nancy (1999). The Dappled World: A Study of the Boundaries of Science.
Cambridge: Cambridge University Press.
Cat, Jordi (2007). “Scientific Unity,” Stanford Online Encyclopedia of Philosophy.
http://plato.stanford.edu/entries/scientific-unity/#Dis.
Clapp, Lenny (2001). “Disjunctive Properties: Multiple Realizations,” Journal of
Philosophy 98: 111-136.
30
Corry, Richard (2002). “A Causal-Structural Theory of Empirical Knowledge,” Ph.D.
thesis, Indiana University.
Corry, Richard (forthcoming). “Scientific Analysis and Causal Influence,” in Handfield
forthcoming.
Crisp, Thomas and Ted Warfield (2001). "Kim's Master Argument: A Critical Notice of
Mind in a Physical World," Noûs 35: 304-16.
Dupré, John (1993). The Disorder of Things: Metaphysical Foundations of the Disunity
of Science. Cambridge, MA: Harvard University Press.
Dupré, John (2001). Human Nature and the Limits of Science. Oxford: Oxford University
Press.
Fodor, Jerry (1974). "Special Sciences," Synthese 28: 97-115.
Gillett, Carl (2002). “The Dimensions of Realization: A Critique of the Standard View,”
Analysis 62: 316-323.
Gillett, Carl (2003a). “Strong Emergence as a Defense of Non-Reductive Physicalism: A
Physicalist Metaphysics for ‘Downward’ Determination,” Principia 6: 83-114.
Gillett, Carl (2003b). “Non-Reductive Realization and Non-Reductive Identity: What
Physicalism does not Entail,” in Walter and Heckmann 2003: 31-57.
Gillett, Carl (2006a). “The Hidden Battles over Emergence,” in P. Clayton, ed., Oxford
Handbook of Religion and Science. Oxford: Oxford University Press, 801-818.
Gillett, Carl (2006b). “Samuel Alexander’s Emergentism: Or, Higher Causation for
Physicalists,” Synthese 153: 261-296.
Handfield, Toby, ed. (forthcoming). Dispositions and Causes. Oxford: Oxford University
Press.
31
Heil, John (2000). From an Ontological Point of View. Oxford: Oxford University Press.
Heil, John (2003). “Multiply Realized Properties,” in Walter and Heckmann 2003: 11-30.
Heil, John and Al Mele, eds. (1993). Mental Causation. Oxford: Clarendon Press.
Hendry, Robin (2006). “Is there downward causation in chemistry?” in D. Baird, L.
McIntyre and E. Scerri, eds., Philosophy of Chemistry: Synthesis of a New
Discipline, Boston Studies in the Philosophy of Science Volume 242 (Springer),
173-189.
Kim, Jaegwon (1993). Supervenience and Mind. Cambridge, MA: MIT Press.
Kim, Jaegwon (1997). "Does the Problem of Mental Causation Generalize?" Proceedings
of the Aristotelian Society 97: 281-97.
Kim, Jaegwon (1998). Mind in a Physical World. Cambridge: Cambridge University
Press.
Kim, Jaegwon (1999). "Supervenient Properties and Micro-based Properties: A Reply to
Noordhof," Proceedings of the Aristotelian Society 99: 115-18.
Kim, Jaegwon (2002). “Responses,” Philosophy and Phenomenological Research 65:
671-680.
Kim, Jaegwon (2003). "Blocking Causal Drainage and Other Maintenance Chores with
Mental Causation," Philosophy and Phenomenological Research 67: 151-76.
Kim, Jaegwon (2005). Physicalism, or Something Near Enough. Princeton, NJ:
Princeton University Press.
Laughlin, R.B., D. Pines, J. Schmalian, B. Stojkovic, and P. Wolynes (2000). “The
middle way,” Proceedings of the National Academy of Sciences 97 (Jan.4): 32-37.
32
Lewis, David (1980). "Mad Pain and Martian Pain," in N. Block, ed., Readings in the
Philosophy of Psychology, Vol.I. Cambridge, MA: Harvard University Press, 216-
22.
Lewis, David (1999). “New Work for a Theory of Universals,” in Papers in Metaphysics
and Epistemology. Cambridge: Cambridge University Press, 8-55.
Loewer, Barry (2001). "Review of J. Kim, Mind in a Physical World," Journal of
Philosophy 98: 315-324.
Mumford, Stephen (2004). Laws in Nature. New York, NY: Routledge.
O’Connor, Timothy (forthcoming). “Agent-Causal Power,” in Handfield forthcoming.
O'Connor, Timothy and Kevin Kimble (forthcoming). "The Argument from
Consciousness Revisited," in Paul Pistone and Quentin Smith, eds., Theism And
Naturalism: New Essays. Oxford: Oxford University Press.
O’Connor, Timothy and H.Y. Wong (2002). "Emergent Properties," Stanford Online
Encyclopedia of Philosophy. http://plato.stanford.edu/entries/properties-emergent/.
O’Connor, Timothy and H.Y. Wong (2005). "The Metaphysics of Emergence," Noûs 39:
659-679.
Pereboom, Derk (2002). “Robust Non-Reductive Materialism,” Journal of Philosophy 99:
499-531.
Pereboom, Derk and Hilary Kornblith (1991). “The Metaphysics of Irreducibility,”
Philosophical Studies 63: 125-145.
Rosenberg, Alexander (1994). Instrumental Biology or the Disunity of Science. Chicago,
IL: University of Chicago Press.
33
Ross, David and David Spurrett (2004). “What to Say to s Skeptical Metaphysician: A
Defense Manual for Cognitive and Behavioral Scientists,” Behavioral and Brain
Sciences 27: 603-647.
Shoemaker, Sydney (1980). “Causality and Properties,” in P. Van Inwagen, ed., Time
and Cause. Dordrecht: Reidel, 109-135.
Shoemaker, Sydney (1998). “Causal and Metaphysical Necessity,” Pacific Philosophical
Quarterly 79: 59-77.
Shoemaker, Sydney (2001). “Realization and Mental Causation,” in C. Gillett and B.
Loewer, eds., Physicalism and its Discontents. Cambridge: Cambridge University
Press, 74-98.
Shoemaker, Sydney (2002). “Kim on Emergence,” Philosophical Studies 108: 53-63.
Shoemaker, Sydney (2007). Physical Realization. Oxford: Oxford University Press.
Tooley, Michael (1993). “Causation: Reductionism versus Realism,” in E. Sosa and M.
Tooley, eds., Causation. Oxford: Oxford University Press, 172-192.
Tooley, Michael (1987). Causation: A Realist Approach. Oxford: Oxford University
Press.
van Fraassen, Bas (1988). Laws and Symmetry. Oxford: Clarendon Press.
van Gulick, Robert (1993). “Who’s in Charge Here? And Who’s Doing All the Work?”
in Heil and Mele 1993: 233-256.
Walter, Sven, and Heinz-Dieter Heckmann, eds. (2003). Physicalism and Mental
Causation: The Metaphysics of Mind and Action. Exeter: Imprint Academic.
Woodward, James (2003). Making Things Happen: A Theory of Causal Explanation.
Oxford: Oxford University Press.
34
Yablo, Stephen (1992). “Mental Causation,” The Philosophical Review 101: 245-280.
i While Kim does not officially endorse a causal powers metaphysic, he has noted his reliance upon a certain view of causation in making his case, a view that, on the surface, bears some similarity to the view we endorse. See for example his 1998: 45-56; 2002: 674-675; and 2005: 17-18, 30, 45, and 47n12. However, there is some ambiguity in the way Kim uses the term “cause.” See his 1998: 43, and 2005: 20 and 41. And see especially his 2002: 677, where he speculates that causality may supervene on fundamental laws (and, perhaps, initial conditions), or that it could “emerge” at macro-levels (“Could Hume be right about fundamental physics but wrong about macro-objects and events?”), or that it might be “implemented” or “realized” by something more basic, like energy flow or momentum transfer. ii One might hold to a philosophical view leading one to insist that in certain cases, it is indeed the object that exercises the power, and not the event of the object’s having the property/causal power. Such is the claim of the agent causationist, e.g., with respect to the forming of a free decision. But this is a substantive and controversial thesis, not a spelling out for one sort of case what is common to every case of causation. (For a discussion of the relationship of agent causation to the more usual ‘event causation’ within a causal powers metaphysics, see Timothy O’Connor, “Agent-Causal Power,” forthcoming in Toby Handfield, ed., Dispositions and Causes, Oxford University Press. iii We are influenced here by Richard Corry. See “A Causal-Structural Theory of Empirical Knowledge” (Ph.D. thesis, Indiana University, 2002) and “Scientific Analysis and Causal Influence,” forthcoming in the volume cited in the previous note. iv We are aware that not everyone shares our temperament with respect to explanation, so not everyone will share our disappointment. For more on causal explanation, to include thorough discussion of extant theories of causal explanation, see Woodward 2003. v See van Fraassen 1988: Ch.5, and Lewis 1999: 40. vi Mumford 2004 includes an extended discussion of this problem for the Armstrongian approach. See pp. 99-103 and 148-149. vii See Kim 1993, 1997, 1998, 1999, 2003, 2005. viii See for example the variation among Fodor 1974; Pereboom and Kornblith 1991 and Pereboom 2002; Shoemaker 2001 and 2007; and Gillett 2002. ix We will ignore the complication of indeterministic causation, which would require us to formulate the completeness thesis in terms of fixing the chances of the effect. x We will, for the sake of convenience, continue to refer only to P, the single realizer of M, though it should be understood that on some accounts of realization M may be realized by multiple properties (“the P’s,” say) each time it is instanced. As we'll see below, Gillett 2002 is one such account. xi For discussion, see Baker 1993; Burge 1993; van Gulick 1993; Kim 1996, 1997, 1999, 2003, 2005; Block 2003; Ross and Spurrett 2004. xii See Fodor 1974; Dupré 1993; and Rosenberg 1994.
35
xiii See Cartwright 1999 and Dupré 1993 and 2001. And for a powerful challenge to the case for completeness in the special scientific domain of chemistry in which it is widely thought to be most secure, see Hendry 2006. xiv On this point, see Heil 2000: Ch.2-7. xv For key passages in support of this interpretation, see Shoemaker 2007: section II of chapter 2 (especially 17-18), 55-56, and section V of chapter 4 (especially 79 and 82). See also the remark that Lenny Clapp has proposed a view similar to his own in Shoemaker 2001: 93n3 and 2007: 11. (In his 2001, Clapp is explicit in his construal of realized properties as disjunctive properties with their disjuncts as the realizers.) xvi Shoemaker 2001: 81. See Yablo 1992. xvii See Shoemaker 2001: 81 and 2007: 13-14. xviii A bolstering consideration comes from certain indeterministic scenarios. We take it to be evident that, assuming the causal completeness of physics, the chance of E given M cannot be greater than the chance of E given a total physical cause (here, our P). But there seems to be no reason to think that it cannot be less. Now consider a case where Pr (E/M) is significantly less than Pr(E/P). Surely in such a case, where E in fact occurs, it is highly implausible to insist nevertheless that M, not P, is the cause of E. While this is a special case, if our conclusion from it is accepted, it indicates further that there is something wrong about Shoemaker’s method for assigning causes. xix We’ve recently come across a paper where a similar interpretation of Shoemaker is entertained. See Heil 2003: 24. xx See the discussion in O'Connor and Wong 2005. xxi See his 2006b: 275, 280, 281, 282. See also Gillett 2002 for the proposal and defense of his preferred account of realization. xxii Gillett 2003a: 102; Gillett 2006b: 274, 281, 282, 287. xxiii Gillett 2003a. xxiv Gillett buttresses his claims about a non-causal determination relation by comparing the case of emergent properties (like H) to what he counts as other cases where there is non-causal determination. Such cases include parts-wholes, realization, constitution, and properties that contribute conditional powers. See Gillett 2003a: 109; 2006a: 6-7; 2006b: 268. Though more deserves to be said in response, we’ll here say only that a causal powers theorist is under no obligation to accept, and may have good reason to reject, each of the four additional candidates for non-causal determination relations Gillett proposes. The grounds in favor of this response are, in brief, that (i) questions about which properties count as causally efficacious (in a causal nonreductionist sense) ought to be settled prior to any commitment concerning the first three proposed relations, and (ii) a causal powers theorist inclined toward “sparseness” with respect to properties will reject the sort of conditional power-conferring properties Gillett invokes (in his 2003a: 101 and 2006b: 279-280, 285-286) as the relata of the fourth proposed relation. xxv Gillett anticipates this analogy by deeming spatial relations entities that (if they exist) “do not contribute powers themselves” but “may still determine the contributions of powers to individuals by other properties and relations” (2003b: 35). xxvi Concepts of emergence have a long history—one need only consider Aristotle’s notion of irreducible substantial forms. Their coherence is also a matter of controversy. For an attempt to sort out the different ideas that have carried this label, see O’Connor
36
and Wong 2002. And for a detailed exposition and defense of the notion we rely on in the text, see O’Connor and Wong 2005. xxvii For numerous examples of such phenomena, see R.B. Laughlin, et al. 2000. xxviii This is not to concede that it is ipso facto a theoretical virtue for a metaphysic that it entails greater unity in nature, nor that it is ipso facto a theoretical vice if the converse is true. The issue of the unity of nature, and the related issue of unity in science, are deep and complex. Our point in the text is that there is a kind of unity in nature if the emergentist account we have proposed is correct. For more on the topics of unity in science and nature, see Cat 2007. xxix For argument on this point, see Timothy O'Connor and Kevin Kimble, "The Argument from Consciousness Revisited," in Paul Pistone and Quentin Smith, eds., Theism And Naturalism: New Essays (Oxford University Press, forthcoming).